Security audit

私人AI助理选型与试用助手

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed assistant-selection workflow with an optional AI-HIVE connection and local result checker; I found no hidden or destructive behavior.

Before installing, be aware this is primarily a Chinese-language workflow and may optionally connect to AI-HIVE through web login or MCP/OAuth. Keep credentials out of chat, authorize uploads or paid actions only for a clearly defined task and budget, and review generated scorecards and external receipts before relying on them.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Tp4

High

Category: MCP Tool Poisoning
Confidence: 99% confidence
Finding: The declared description promises a user-facing assistant-selection helper that analyzes preferences and produces trial questionnaires, CSV scorecards, and selection advice. The supplied code does none of that. Its primary purpose is to validate a saved result schema/contract and local artifact integrity, including status consistency, authorization references for mutating actions, provenance fields, and SHA-256 checks. While it does not perform dangerous undeclared external actions, its core functionality is materially different and unrelated to the declared product behavior. Therefore this is a clear description-behavior mismatch.
Content: No source excerpt is available for this finding.

Undeclared Tool Scope

Medium

Category: MCP Least Privilege
Confidence: 70% confidence
Finding: Without declared permissions the skill's intent is opaque and cannot be validated.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: This markdown file presents all operational instructions exclusively in Chinese, which can amount to a language policy violation when no user opt-in or justified locale constraint is provided. The document does not state that the skill is region-specific or offer an alternative language.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The natural-language note is written only in Chinese: '空值供用户按实际任务填写，不填密码、Token、验证码、Cookie或完整证件信息。' For a general intake file with no stated regional or compliance scope, this imposes a language expectation without user opt-in or an explicit locale justification.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Low

Category: Not specified by scanner
Confidence: 81% confidence
Finding: The docstring presents the tool as if it performs no actions, yet the implementation reads result files, traverses artifact paths under a supplied root, checks file metadata, and hashes file contents. While these are local-only operations and not remote actions, the wording "executes actions" is broader than the actual behavior and can mislead about side effects and file access.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.